1. Field of the Invention
The present invention relates to an image processing apparatus, an image forming apparatus, an image processing method, and a computer readable recording medium, each capable of converting image data inputted thereto into image data with a predetermined gray level level.
2. Description of the Related Art
In recent years, digitalization of office automation (OA) equipment has rapidly progressed, and a demand for color image output has increased, which has caused image output apparatuses such as an electrophotographic digital color copier or an inkjet or thermal transfer color printer to be in widespread use. For example, image data inputted from an image input apparatus such as a digital camera or a scanner or image data created with a computer is outputted using such image output apparatuses. Since the number of gray levels of an image, which can be outputted from the image output apparatus, is smaller than the number of gray levels of image data inputted thereto in most cases, in such an image output apparatus, a pseudo tone reproduction process is performed on the inputted image data before outputting the image.
As one of the tone reproduction processes, an error diffusion method is cited. Since a fixed error diffusion coefficient is applied to all of the pixels in the error diffusion method, an accumulative error becomes periodical, resulting in textures or worms.
In order for preventing generation of such textures, there is proposed a technology for switching a position of the error diffusion coefficient at random by switching a position of a weighting coefficient with a random number (refer to Toshiharu Kurosawa, “Output image characteristic in Modified Error Diffusion (MED) method,” The journal of The Institute of Image Electronics Engineers of Japan, 1991, vol. 20, No. 5, pp. 443-449, for example).
Further, as a process having a property of achieving substantially the same image quality as the error diffusion method even with a dither method, there is cited a blue noise mask method. The blue noise mask method has the same processing method as the dither method except the threshold mask, which is large and has a similar arrangement of the threshold values to the dot profile of the error diffusion method. Therefore, it requires only smaller size of hardware than that of the error diffusion method for performing the pseudo tone reproduction process. In inkjet printers using dark and light inks, the error diffusion method is applied to images reproduced with the light ink while the blue noise mask method is applied to images reproduced by the dark ink because of a problem of affinity of the dot arrangement (refer to Tsuyoshi Makita, “High-Quality Picture Technologies in Inkjet Printers,” The journal of The Institute of Image Electronics Engineers of Japan, 2001, vol. 40, No. 3, pp. 237-243, for example).
In the case in which all of the pixels are processed with the blue noise mask method, there arises a problem that the peripheries of characters and line drawings become awkwardly spattering in the output, or that an awkward granular output is caused in a halftone dot area by interference with the original halftone pattern, although different form the moire occurred in the dither method, unless the halftone dot area is sufficiently smoothed in the former process.
Further, in the process using the error diffusion method, it is required for preventing generation of the textures to switch the position of the error diffusion coefficient based on a random number as described above or to diffuse the quantization error two lines ahead thereof. Thus, there arises a problem that in the error diffusion method which requires an already large-scale apparatus, growth in size of the processing circuit occurs, or a large capacity line buffer for storing the quantization error or the accumulated error becomes necessary.